Laser intrusion detection method and device based on weather conditions, equipment and medium

Abstract

The application relates to the technical field of laser intrusion detection, and discloses a laser intrusion detection method, device and equipment based on meteorological conditions, which comprises the following steps: collecting meteorological data and laser signal data, and calculating a meteorological self-adaptive fusion coefficient according to the meteorological data; calculating a laser signal quality evaluation parameter according to the laser signal data; performing channel self-attention calculation according to the laser signal data to obtain channel attention features; inputting the laser signal quality evaluation parameter, the meteorological self-adaptive fusion coefficient and the channel attention features into a pre-constructed numerical weather model to output a laser intrusion probability, and comparing the laser intrusion probability with a preset intrusion probability threshold to determine whether intrusion occurs. The application can reduce noise alarm and false alarm rate under complex meteorological conditions, and improve the intrusion detection accuracy based on meteorological self-adaptation.

Description

Technical Field

[0001] This invention relates to the field of laser intrusion detection technology, and in particular to a laser intrusion detection method, apparatus and equipment based on meteorological conditions. Background Technology

[0002] A laser intrusion detector is a beam-type active intrusion detection device consisting of a transmitter, a receiver, and a backend server. After installing the transmitter and receiver at opposite ends of the protected area, multiple parallel laser beams form an invisible laser fence between the two ends. The transmitter continuously emits laser pulses towards the receiver, which is hundreds of meters away. Upon receiving the laser signal, the receiver forms a complete optical path with the transmitter. When an intruder crosses the protected area, the optical path is physically blocked, and the receiver fails to detect the expected laser signal. The system then triggers an alarm and reports the alarm signal to the server via RS485 or Ethernet, triggering the activation of security equipment such as audible and visual alarms, electronic maps, video surveillance systems, and lighting systems. Simultaneously, the receiver starts a timing function every time it is powered on, calculating resource load in real time, detecting light source intensity, calculating light source lifespan, and reporting this data to the server. It also receives configuration information from the server, including sensitivity adjustments, alarm strategies, and weather handling settings.

[0003] However, traditional laser intrusion detection systems have a fundamental limitation: their core control parameters, such as laser emission power, receiver gain, signal detection threshold, and filtering parameters, are fixed after installation and commissioning. The system cannot sense changes in the external environment throughout its lifespan, let alone adapt to them. While attempts have been made to add independent weather stations alongside the laser beam system, this simple physical juxtaposition has not achieved deep data fusion and intelligent linkage. There is a lack of real-time, adaptive decision-making and control closed loop between meteorological data and laser control parameters.

[0004] When encountering complex weather conditions such as rain, fog, snow, flying insects, and drastic changes in lighting, the system frequently generates noise alarms and false alarms, resulting in a severe decline in detection performance. During long-distance transmission, the laser signal quality is degraded due to factors such as rain and fog scattering and atmospheric attenuation, making it difficult for the system to accurately distinguish between genuine intrusion behavior and environmental interference.

[0005] In summary, how to reduce noise alarms and false alarm rates under complex weather conditions, and improve the accuracy of intrusion detection based on weather adaptation, has become an important technical problem that urgently needs to be solved in the field of laser intrusion detection technology. Summary of the Invention

[0006] This invention provides a laser intrusion detection method, device, electronic device, and computer-readable storage medium based on meteorological conditions, which can reduce noise alarm and false alarm rates under complex meteorological conditions and improve intrusion detection accuracy based on meteorological adaptation.

[0007] To achieve the above objectives, the present invention provides a laser intrusion detection method based on meteorological conditions, comprising: Collect meteorological data and laser signal data, and calculate the meteorological adaptive fusion coefficient based on the meteorological data; Calculate laser signal quality evaluation parameters based on the laser signal data; Channel self-attention is calculated based on the laser signal data to obtain channel attention characteristics; The laser signal quality evaluation parameters, the meteorological adaptive fusion coefficients, and the channel attention features are input into a pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is compared with a preset intrusion probability threshold to determine whether an intrusion has occurred.

[0008] Optionally, the laser signal data is acquired through a laser transmitter and a laser receiver; The meteorological data collection methods include: real-time data collection through external meteorological sensors, real-time weather forecast data acquisition through meteorological service APIs, and manually input weather data acquisition through operation and maintenance interfaces.

[0009] Optionally, calculating the laser signal quality evaluation parameters based on the laser signal data includes: The laser signal quality evaluation parameters include pulse integrity ratio, response time jitter, and waveform similarity. The pulse integrity ratio is obtained by calculating the ratio of the total number of transmitted pulses to the number of effective received pulses in the laser signal data. The response time jitter is obtained by calculating the difference based on the arrival time of the emitted pulse in the laser signal data; The similarity is calculated based on the transmitted and received waveforms in the laser signal data to obtain the waveform similarity.

[0010] Optionally, the step of performing channel self-attention calculation based on the laser signal data to obtain channel attention features includes: The channel attention features include the average occlusion level feature corresponding to average pooling and the peak occlusion intensity feature corresponding to max pooling; Global average pooling and global max pooling are performed on the received signal feature map in the laser signal data respectively to obtain the corresponding average pooling signal features and max pooling signal features; The average pooling signal features and the max pooling signal features are input into a multilayer perceptron using the following formulas to generate corresponding average occlusion level features and peak occlusion intensity features:

[0011]

[0012] in, This is represented as the average occlusion level feature. Represented as peak occlusion intensity characteristics, Represented as a received signal feature map, This is represented as global average pooling. This is represented as global max pooling. This is represented as a multilayer perceptron.

[0013] Optionally, the meteorological data is multi-dimensional vector data, including dimensions corresponding to rainfall, snowfall, haze, and light intensity; After collecting meteorological data, the process also includes vectorizing the meteorological data to obtain meteorological condition vectors.

[0014] Optionally, calculating the meteorological adaptive fusion coefficient based on the meteorological data includes: The meteorological condition vector corresponding to the meteorological data is activated using the following formula to obtain the meteorological adaptive fusion coefficient;

[0015] in, Represented as the meteorological adaptive fusion coefficient, Represented as a meteorological condition vector, Represented as the transpose symbol, Represented as preset learnable weights, This is represented as a preset learnable bias term. Use the Sigmoid activation function; When the weather adaptive fusion coefficient approaches 1, it indicates good weather; when the weather adaptive fusion coefficient approaches 0, it indicates bad weather.

[0016] Optionally, the numerical weather model is represented as:

[0017] in, This is represented as a received signal feature map in the laser signal data. This is represented as the meteorological condition vector corresponding to the meteorological data. This is expressed as the probability of laser intrusion. These are respectively represented as the average occlusion level feature and the peak occlusion intensity feature in the meteorological adaptive fusion coefficient. Represented as the meteorological adaptive fusion coefficient, These are respectively represented by the pulse integrity ratio, response time jitter, and waveform similarity in the laser signal quality evaluation parameters. These represent the learnable scalar weights corresponding to pulse integrity ratio, response time jitter, and waveform similarity, respectively.

[0018] To address the aforementioned problems, the present invention also provides a laser intrusion detection device based on meteorological conditions, the device comprising: The data acquisition module is used to collect meteorological data and laser signal data; The meteorological adaptive fusion coefficient calculation module is used to calculate the meteorological adaptive fusion coefficient based on the meteorological data; A laser signal quality evaluation parameter calculation module is used to calculate laser signal quality evaluation parameters based on the laser signal data. The channel attention feature calculation module is used to perform channel self-attention calculation based on the laser signal data to obtain channel attention features; The intrusion probability calculation and judgment module is used to input the laser signal quality evaluation parameters, the meteorological adaptive fusion coefficient and the channel attention features into a pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is used to compare with a preset intrusion probability threshold to determine whether an intrusion has occurred.

[0019] To address the above problems, the present invention also provides an electronic device, the electronic device comprising: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the above-described laser intrusion detection method based on meteorological conditions.

[0020] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one computer program, which is executed by a processor in an electronic device to implement the aforementioned laser intrusion detection method based on meteorological conditions.

[0021] This invention significantly reduces noise alarms and false alarms caused by environmental interference such as rain, fog, snow, flying insects, and strong light by deeply integrating the channel attention mechanism with physical logic parameters. The channel attention module further automatically redistributes attention weights at the feature level, enabling it to accurately distinguish between real intrusions and environmental interference even when meteorological sensor classification is not refined enough, weather changes are drastic, or local environmental conditions are complex. In terms of adaptive capability in complex meteorological environments, the weights of deep learning features and physical logic terms are dynamically adjusted through meteorological adaptive fusion coefficients. It maintains high sensitivity in clear weather and automatically switches to a highly robust mode in severe weather without human intervention, achieving intelligent laser intrusion detection under all weather conditions. Attached Figure Description

[0022] Figure 1 A schematic flowchart of a laser intrusion detection method based on meteorological conditions provided in an embodiment of the present invention; Figure 2 A functional block diagram of a laser intrusion detection device based on meteorological conditions provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device for implementing the laser intrusion detection method based on meteorological conditions, according to an embodiment of the present invention.

[0023] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0024] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0025] This application provides a meteorological-based laser intrusion detection method applied to active intrusion detection scenarios in perimeter security. It introduces a triple-fusion self-learning mechanism combining data-driven, physical prior, and meteorological adaptive approaches. The core of this mechanism lies in breaking away from the limitations of traditional systems that process laser signals in isolation, using multi-source meteorological data as a dynamic adjustment lever. On one hand, it utilizes the channel attention mechanism in deep learning to extract microscopic signal features as the basis for data-driven operations; on the other hand, it delves into the physical channel state of the laser, extracting parameters such as pulse integrity and time jitter as physical prior knowledge. Most importantly, by calculating the meteorological adaptive fusion coefficient, the system can dynamically adjust the fusion ratio of data-driven features and physical prior logic within a numerical weather model. Under favorable weather conditions, it relies on data-driven approaches to maintain high sensitivity, while under adverse weather conditions, it retreats to physical priors to ensure high robustness. Finally, by establishing a complete closed-loop workflow from meteorological data input, intelligent analysis and calculation to batch parameter configuration, it achieves optimal detection performance in all weather conditions, effectively reducing noise alarm rates.

[0026] Reference Figure 1 The diagram shown is a schematic flowchart of a laser intrusion detection method based on meteorological conditions according to an embodiment of the present invention. In this embodiment, the laser intrusion detection method based on meteorological conditions includes: S1. Collect meteorological data and laser signal data, and calculate the meteorological adaptive fusion coefficient based on the meteorological data.

[0027] In a practical application scenario of this invention, meteorological data is collected in real time through three input methods to ensure data reliability: First, data is collected in real time through external professional meteorological sensors deployed on-site. These sensors may include anemometers, rain gauges, visibility meters, and illuminance sensors, used to collect local micro-meteorological data. Second, national or commercial meteorological service APIs are called to obtain accurate real-time weather data and short-term forecasts for the area where the equipment is located. Third, a management interface is provided, allowing maintenance personnel to manually input or select the current weather conditions based on actual observations. These three methods complement and cross-validate each other to ensure the reliability and comprehensiveness of the meteorological data.

[0028] Furthermore, the collected meteorological data can be vectorized to form a meteorological condition vector. This vector is multidimensional data, containing at least four dimensions: rainfall intensity, snowfall intensity, haze concentration, and light intensity. Each component is normalized to the range of 0 to 1, representing the severity of the corresponding meteorological element.

[0029] In this embodiment of the invention, the original signal data is obtained through the collaborative work of the laser transmitter and receiver. The laser signal data includes the total number of transmitted pulses, the number of effective received pulses, the arrival time of the transmitted pulses, the transmitted waveform, the received waveform, and the received signal feature map, etc.

[0030] Specifically, the transmitter sends a modulated laser pulse sequence to the receiver and records the total number of transmitted pulses. The receiver's photodetector converts the received optical signal into an electrical signal, which, after signal conditioning and analog-to-digital conversion, is then digitized by a microcontroller. The receiver counts the number of valid received pulses, records the arrival time of each pulse, and performs intensive sampling of the received waveform using a high-speed ADC to generate a digitized received signal feature map. This feature map is a two-dimensional data matrix—rows correspond to different beam channels (e.g., 6 laser beams = 6 channels), columns correspond to the continuously sampled time series, and the values ​​in the matrix represent the attenuation rate of the received light intensity of each beam at each moment.

[0031] In this embodiment of the invention, calculating the meteorological adaptive fusion coefficient based on the meteorological data includes: The meteorological condition vector corresponding to the meteorological data is activated using the following formula to obtain the meteorological adaptive fusion coefficient;

[0032] in, Represented as the meteorological adaptive fusion coefficient, Represented as a meteorological condition vector, Represented as the transpose symbol, Represented as preset learnable weights, This is represented as a preset learnable bias term. Use the Sigmoid activation function; When the weather adaptive fusion coefficient approaches 1, it indicates good weather; when the weather adaptive fusion coefficient approaches 0, it indicates bad weather.

[0033] Specifically, when the weather is clear (rain, snow, and fog components in W are 0, and sunlight is normal), When the value is large and negative, γ approaches 0, at which point deep learning features (channel attention features) dominate; when the weather is severe (heavy rain, dense fog, etc.), When γ is a large positive value, it approaches 1. At this point, the weight of the physical logic term is increased, and the deep learning features are calibrated using the principles of laser physics.

[0034] S2. Calculate the laser signal quality evaluation parameters based on the laser signal data.

[0035] In this embodiment of the invention, after obtaining the laser signal data, it is processed in parallel from at least two dimensions: first, extracting laser signal quality evaluation parameters (i.e., physical logic terms), and second, calculating channel attention features (i.e., data-driven features).

[0036] In this embodiment of the invention, calculating the laser signal quality evaluation parameters based on the laser signal data includes: The laser signal quality evaluation parameters include pulse integrity ratio, response time jitter, and waveform similarity. The pulse integrity ratio is obtained by calculating the ratio of the total number of transmitted pulses to the number of effective received pulses in the laser signal data. The response time jitter is obtained by calculating the difference based on the arrival time of the emitted pulse in the laser signal data; The similarity is calculated based on the transmitted and received waveforms in the laser signal data to obtain the waveform similarity.

[0037] Specifically, the pulse integrity ratio is the ratio of the number of normally received pulses to the total number of transmitted pulses. This ratio directly reflects the degree to which the optical path is blocked. When an object passes through the light beam, the number of received pulses decreases, and the pulse integrity ratio decreases accordingly.

[0038] For each successfully received valid optical pulse, the laser receiver records the arrival time of the transmitted pulse. Based on the set of arrival timestamps of all pulses received by the receiver within a statistical window (e.g., with a transmission time reference of 1 second), the maximum and minimum arrival times of the transmitted pulses can be obtained, i.e., the maximum and minimum response times. By calculating their difference, multipath interference and signal stability can be reflected. When the response time jitter is less than a preset jitter threshold, it indicates that the signal path is single and stable, and is in an ideal state. The threshold value ranges from 1ms to 10ms. When the response time jitter is greater than the threshold, it indicates the presence of reflection, scattering, or interference paths, such as scattering caused by environmental factors like rain, fog, and dust.

[0039] The waveform similarity can be obtained by using the Pearson correlation coefficient between the transmitted and received signal waveforms. This coefficient reflects the signal quality and the degree of distortion. When the waveform similarity is close to 1, it indicates that the transmitted and received waveforms are highly consistent and the signal quality is clean. When the waveform similarity is much less than 1, it indicates that the signal is distorted due to strong light interference, device aging, or contamination.

[0040] S3. Perform channel self-attention calculation based on the laser signal data to obtain channel attention characteristics.

[0041] In this embodiment of the invention, the step of performing channel self-attention calculation based on the laser signal data to obtain channel attention features includes: The channel attention features include the average occlusion level feature corresponding to average pooling and the peak occlusion intensity feature corresponding to max pooling; Global average pooling and global max pooling are performed on the received signal feature map in the laser signal data respectively to obtain the corresponding average pooling signal features and max pooling signal features; The average pooling signal features and the max pooling signal features are input into a multilayer perceptron using the following formulas to generate corresponding average occlusion level features and peak occlusion intensity features:

[0042]

[0043] in, This is represented as the average occlusion level feature. Represented as peak occlusion intensity characteristics, Represented as a received signal feature map, This is represented as global average pooling. This is represented as global max pooling. This is represented as a multilayer perceptron.

[0044] Specifically, global average pooling averages all sampled values ​​of each laser beam over time to reflect the average occlusion level of the beam, which is very effective for identifying large-area rain and fog attenuation; global maximum pooling selects the lowest value from the time series of each laser beam, i.e. the value at the moment of most severe occlusion, to reflect the peak occlusion intensity of the beam, which has extremely high sensitivity for identifying small objects (such as birds or fallen leaves) or human limbs that pass quickly through the beam.

[0045] The average pooling and max pooling results are fed into a multilayer perceptron (MLP) to generate two feature scalars: average occlusion level feature and peak occlusion intensity feature. These two features together constitute the channel attention feature. The average occlusion level feature focuses on the persistence of occlusion, while the peak occlusion intensity feature focuses on the severity of occlusion.

[0046] The multilayer perceptron learned to cross-compare the data patterns of each beam through training. When it found that a beam was continuously and extensively blocked while other beams were normal, it was judged as a characteristic pattern of suspected human intrusion. When all beams showed random and brief fluctuations, it was judged as environmental interference such as raindrops or flying insects.

[0047] This invention introduces channel self-attention computation, enabling it to automatically learn and evaluate the importance weight of each feature channel. Channels containing key patterns of human intrusion are assigned higher weights to amplify these features; conversely, channels containing only raindrop noise or insect interference are assigned extremely low weights to suppress this noise. This adaptive redistribution at the micro-feature level can still isolate surface interference and avoid misjudgments at the micro-feature level, even when meteorological sensor classification is not precise enough or the local environment is extremely complex.

[0048] S4. Input the laser signal quality evaluation parameters, the meteorological adaptive fusion coefficient, and the channel attention features into the pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is used to compare with a preset intrusion probability threshold to determine whether an intrusion has occurred.

[0049] In this embodiment of the invention, the numerical weather model is represented as:

[0050] in, This is represented as a received signal feature map in the laser signal data. This is represented as the meteorological condition vector corresponding to the meteorological data. This is expressed as the probability of laser intrusion. These are respectively represented as the average occlusion level feature and the peak occlusion intensity feature in the meteorological adaptive fusion coefficient. Represented as the meteorological adaptive fusion coefficient, These are respectively represented by the pulse integrity ratio, response time jitter, and waveform similarity in the laser signal quality evaluation parameters. These represent the learnable scalar weights corresponding to pulse integrity ratio, response time jitter, and waveform similarity, respectively.

[0051] The outer Sigmoid function compresses the weighted sum within the parentheses to the range of 0 to 1, so that the closer the output value is to 1, the higher the confidence level of the system in determining it as a real intrusion, and the closer it is to 0, the higher the confidence level of the system in determining it as safe.

[0052] This numerical weather model achieves a triple fusion of data-driven, physical prior, and meteorological adaptive components: the data-driven part (A+M) adaptively learns complex patterns from laser signal data; the physical prior part (ω1S+ω2D+ω3R) injects domain knowledge of laser beams to improve interpretability and robustness; and the meteorological adaptive part (γ) dynamically adjusts the fusion ratio of the two components based on real-time weather.

[0053] The final output laser intrusion probability M(F,W) is compared with a preset intrusion probability threshold. When M(F,W) is greater than the threshold, an intrusion is detected and an alarm is triggered; otherwise, silence is maintained. This threshold can be dynamically adjusted according to weather conditions. In severe weather, the threshold can be appropriately increased to reduce the false alarm rate, while in clear weather, the threshold can be appropriately decreased to maintain high sensitivity.

[0054] In a practical application scenario of this invention, after determining the intrusion probability, the system further performs batch configuration parameter operations. Based on real-time meteorological environment and physical parameter status, the system automatically calculates and batch-distributes core parameters to all laser transmitters and receivers on site, including but not limited to: laser emission power (appropriately increased in rainy or foggy weather to compensate for atmospheric attenuation), receiver gain (reduced under strong light to prevent saturation), signal filtering intensity (enhanced filtering in windy weather to suppress vibration noise), and alarm confirmation time (appropriately extended in severe weather to avoid false alarms due to momentary obstruction).

[0055] The above parameters can be distributed in batches via the platform interface, or maintenance personnel can manually adjust them via buttons. In this way, the system realizes a two-level adaptive framework of macro-parameter adaptation (adjusting hardware parameters such as transmit power and receive sensitivity) and micro-feature adaptation (redistributing weights at the feature level by the channel attention mechanism).

[0056] This invention also possesses a self-learning function, capable of periodically collecting historical operational data, including the current meteorological conditions W, laser signal characteristics (A, M), physical parameters (S, D, R), the intrusion probability M(F, W) output by the model, and the final manual calibration result of whether an intrusion actually occurred. Based on this data, the system updates all learnable parameters through optimization algorithms such as gradient descent, including the weights of the MLP in the channel attention module, the calculation parameters (μ and b) of the meteorological adaptive fusion coefficient γ, the learnable scalar weights of the physical parameters (ω1, ω2, ω3), and the mapping parameters of the dynamic threshold, etc.

[0057] For example, if the model outputs an intrusion probability greater than a threshold and triggers an alarm, but no intrusion actually occurs, it indicates a false alarm. The system adjusts the model parameters accordingly to output a lower intrusion probability under the same or similar weather conditions. Through continuous self-learning iteration, the system can continuously adapt to the climate characteristics and field environment of different regions.

[0058] In addition, the system also has sample comparison and pattern recognition capabilities. When laser signal occlusion is detected, the system compares and analyzes the characteristics of the occlusion event with historical samples in the model library to distinguish between normal occlusion and abnormal occlusion.

[0059] For example, obstructions caused by natural phenomena such as falling objects due to typhoons, raindrops, and snowflakes are judged as normal natural phenomena by the system through pattern recognition algorithms and will not trigger an alarm; however, continuous and regular obstructions caused by human intrusion are judged as abnormal obstructions and will trigger an alarm.

[0060] The system can simulate different weather combinations to cover most meteorological conditions. When the actual weather matches the simulated conditions, the equipment parameters can be adjusted in batches manually or automatically, enabling the equipment to adaptively adjust its operating strategy according to meteorological conditions.

[0061] In summary, this invention significantly reduces noise alarms and false alarms caused by environmental interference such as rain, fog, snow, flying insects, and strong light through the deep fusion of channel attention mechanism and physical logic parameters. The channel attention module further automatically reallocates attention weights at the feature level, enabling accurate differentiation between real intrusions and environmental interference even when meteorological sensor classification is not refined enough, weather changes are drastic, or local environmental conditions are complex. In terms of adaptive capability in complex meteorological environments, the weights of deep learning features and physical logic terms are dynamically adjusted through meteorological adaptive fusion coefficients, maintaining high sensitivity in clear weather and automatically switching to a highly robust mode in severe weather without manual intervention, achieving intelligent laser intrusion detection under all weather conditions.

[0062] like Figure 2 The diagram shown is a functional block diagram of a laser intrusion detection device based on meteorological conditions provided in an embodiment of the present invention.

[0063] The meteorological condition-based laser intrusion detection device 100 of this invention can be installed in an electronic device. Depending on the functions implemented, the meteorological condition-based laser intrusion detection device 100 may include a data acquisition module 101, a meteorological adaptive fusion coefficient calculation module 102, a laser signal quality evaluation parameter calculation module 103, a channel attention feature calculation module 104, and an intrusion probability calculation and judgment module 105. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0064] In this embodiment, the functions of each module / unit are as follows: The data acquisition module 101 is used to acquire meteorological data and laser signal data; The meteorological adaptive fusion coefficient calculation module 102 is used to calculate the meteorological adaptive fusion coefficient based on the meteorological data. The laser signal quality evaluation parameter calculation module 103 is used to calculate laser signal quality evaluation parameters based on the laser signal data. The channel attention feature calculation module 104 is used to perform channel self-attention calculation based on the laser signal data to obtain channel attention features; The intrusion probability calculation and judgment module 105 is used to input the laser signal quality evaluation parameters, the meteorological adaptive fusion coefficient and the channel attention feature into a pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is used to compare with a preset intrusion probability threshold to determine whether an intrusion has occurred.

[0065] In detail, each module in the meteorological condition-based laser intrusion detection device 100 described in this embodiment of the invention uses the same technical means as the meteorological condition-based laser intrusion detection method described in the accompanying drawings, and can produce the same technical effect, which will not be repeated here.

[0066] like Figure 3 The diagram shown is a structural schematic of an electronic device for implementing a laser intrusion detection method based on meteorological conditions, according to an embodiment of the present invention.

[0067] The electronic device 200 may include a processor 201, a memory 202, a communication bus 203, and a communication interface 204. It may also include a computer program stored in the memory 202 and capable of running on the processor 201, such as a laser intrusion detection program based on weather conditions.

[0068] In some embodiments, the processor 201 may be composed of an integrated circuit. The processor 201 is the control unit of the electronic device, which connects various components of the electronic device through various interfaces and lines. It performs various functions of the electronic device and processes data by running or executing programs or modules stored in the memory 202 (such as executing a laser intrusion detection program based on weather conditions) and calling data stored in the memory 202.

[0069] The memory 202 includes at least one type of readable storage medium. In other embodiments, the memory 202 may also be an external storage device for the electronic device. The memory 202 may include both internal storage units and external storage devices for the electronic device. The memory 202 can be used not only to store application software and various types of data installed on the electronic device, such as the code of a laser intrusion detection program based on weather conditions, but also to temporarily store data that has been output or will be output.

[0070] The communication bus 203 can be a peripheral component interconnection standard bus or an extended industry standard structure bus, etc. The bus is configured to realize the connection and communication between the memory 202 and at least one processor 201, etc.

[0071] The communication interface 204 is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface.

[0072] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3 The structure shown does not constitute a limitation on the electronic device 200, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0073] For example, although not shown, the electronic device may also include a power source (such as a battery) to power the various components. The electronic device may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described further here.

[0074] It should be understood that the embodiments described are for illustrative purposes only and are not limited to this structure in the scope of the patent application.

[0075] The weather-condition-based laser intrusion detection program stored in the memory 202 of the electronic device 200 is a combination of multiple instructions. When run in the processor 201, it can achieve the following: Collect meteorological data and laser signal data, and calculate the meteorological adaptive fusion coefficient based on the meteorological data; Calculate laser signal quality evaluation parameters based on the laser signal data; Channel self-attention is calculated based on the laser signal data to obtain channel attention characteristics; The laser signal quality evaluation parameters, the meteorological adaptive fusion coefficients, and the channel attention features are input into a pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is compared with a preset intrusion probability threshold to determine whether an intrusion has occurred.

[0076] Specifically, the specific implementation method of the processor 201 of the above instructions can be referred to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

[0077] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following: Collect meteorological data and laser signal data, and calculate the meteorological adaptive fusion coefficient based on the meteorological data; Calculate laser signal quality evaluation parameters based on the laser signal data; Channel self-attention is calculated based on the laser signal data to obtain channel attention characteristics; The laser signal quality evaluation parameters, the meteorological adaptive fusion coefficients, and the channel attention features are input into a pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is compared with a preset intrusion probability threshold to determine whether an intrusion has occurred.

[0078] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0079] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0080] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A laser intrusion detection method based on meteorological conditions, characterized in that, The method includes: Collect meteorological data and laser signal data, and calculate the meteorological adaptive fusion coefficient based on the meteorological data; Calculate laser signal quality evaluation parameters based on the laser signal data; Channel self-attention is calculated based on the laser signal data to obtain channel attention characteristics; The laser signal quality evaluation parameters, the meteorological adaptive fusion coefficients, and the channel attention features are input into a pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is compared with a preset intrusion probability threshold to determine whether an intrusion has occurred.

2. The laser intrusion detection method based on meteorological conditions according to claim 1, characterized in that, The laser signal data is acquired through a laser transmitter and a laser receiver. The meteorological data collection methods include: real-time data collection through external meteorological sensors, real-time weather forecast data acquisition through meteorological service APIs, and manually input weather data acquisition through operation and maintenance interfaces.

3. The laser intrusion detection method based on meteorological conditions as described in claim 1, characterized in that, The calculation of laser signal quality evaluation parameters based on the laser signal data includes: The laser signal quality evaluation parameters include pulse integrity ratio, response time jitter, and waveform similarity. The pulse integrity ratio is obtained by calculating the ratio of the total number of transmitted pulses to the number of effective received pulses in the laser signal data. The response time jitter is obtained by calculating the difference based on the arrival time of the emitted pulse in the laser signal data; The similarity is calculated based on the transmitted and received waveforms in the laser signal data to obtain the waveform similarity.

4. The laser intrusion detection method based on meteorological conditions as described in claim 1, characterized in that, The step of performing channel self-attention calculation based on the laser signal data to obtain channel attention features includes: The channel attention features include the average occlusion level feature corresponding to average pooling and the peak occlusion intensity feature corresponding to max pooling; Global average pooling and global max pooling are performed on the received signal feature map in the laser signal data respectively to obtain the corresponding average pooling signal features and max pooling signal features; The average pooling signal features and the max pooling signal features are input into a multilayer perceptron using the following formulas to generate corresponding average occlusion level features and peak occlusion intensity features: in, This is represented as the average occlusion level feature. Represented as peak occlusion intensity characteristics, Represented as a received signal feature map, This is represented as global average pooling. This is represented as global max pooling. This is represented as a multilayer perceptron.

5. The laser intrusion detection method based on meteorological conditions as described in claim 1, characterized in that, The meteorological data is multidimensional vector data, including dimensions corresponding to rainfall, snowfall, haze, and light intensity; After collecting meteorological data, the process also includes vectorizing the meteorological data to obtain meteorological condition vectors.

6. The laser intrusion detection method based on meteorological conditions as described in claim 5, characterized in that, The calculation of the meteorological adaptive fusion coefficient based on the meteorological data includes: The meteorological condition vector corresponding to the meteorological data is activated using the following formula to obtain the meteorological adaptive fusion coefficient; in, Represented as the meteorological adaptive fusion coefficient, Represented as a meteorological condition vector, Represented as the transpose symbol, Represented as preset learnable weights, This is represented as a preset learnable bias term. Use the Sigmoid activation function; When the weather adaptive fusion coefficient approaches 1, it indicates good weather; when the weather adaptive fusion coefficient approaches 0, it indicates bad weather.

7. The laser intrusion detection method based on meteorological conditions as described in claim 1, characterized in that, The numerical weather model is represented as follows: in, This is represented as a received signal feature map in the laser signal data. This is represented as the meteorological condition vector corresponding to the meteorological data. This is expressed as the probability of laser intrusion. These are respectively represented as the average occlusion level feature and the peak occlusion intensity feature in the meteorological adaptive fusion coefficient. Represented as the meteorological adaptive fusion coefficient, These are respectively represented by the pulse integrity ratio, response time jitter, and waveform similarity in the laser signal quality evaluation parameters. These represent the learnable scalar weights corresponding to pulse integrity ratio, response time jitter, and waveform similarity, respectively.

8. A laser intrusion detection device based on meteorological conditions, characterized in that, The device includes: The data acquisition module is used to collect meteorological data and laser signal data; The meteorological adaptive fusion coefficient calculation module is used to calculate the meteorological adaptive fusion coefficient based on the meteorological data; A laser signal quality evaluation parameter calculation module is used to calculate laser signal quality evaluation parameters based on the laser signal data. The channel attention feature calculation module is used to perform channel self-attention calculation based on the laser signal data to obtain channel attention features; The intrusion probability calculation and judgment module is used to input the laser signal quality evaluation parameters, the meteorological adaptive fusion coefficient and the channel attention features into a pre-constructed numerical weather model to output the laser intrusion probability. The laser intrusion probability is used to compare with a preset intrusion probability threshold to determine whether an intrusion has occurred.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the meteorological condition-based laser intrusion detection method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the laser intrusion detection method based on meteorological conditions as described in any one of claims 1 to 7.

Images